Multidrug Resistance in Staphylococcus aureus Due to ...

Download PDF
2 downloads 0 Views 154KB Size Report
Comment
Aug 24, 2004 - Transport Protein. Glenn W. Kaatz,1,2* Fionnuala McAleese,3 and Susan M. Seo2 ... Mueller-Hinton II broth ... E-mail: gkaatz@juno.com. 1857 ...
ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, May 2005, p. 1857–1864 0066-4804/05/$08.00⫹0 doi:10.1128/AAC.49.5.1857–1864.2005 Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Vol. 49, No. 5

Multidrug Resistance in Staphylococcus aureus Due to Overexpression of a Novel Multidrug and Toxin Extrusion (MATE) Transport Protein Glenn W. Kaatz,1,2* Fionnuala McAleese,3 and Susan M. Seo2 The John D. Dingell Department of Veteran’s Affairs Medical Center1 and The Department of Medicine, Division of Infectious Diseases, Wayne State University School of Medicine,2 Detroit, Michigan, and Wyeth Research, Pearl River, New York3 Received 24 August 2004/Returned for modification 9 October 2004/Accepted 23 January 2005

Efflux is an important mechanism of multidrug resistance (MDR) in bacteria. The multidrug and toxin extrusion (MATE) family is the most recently described group of MDR efflux proteins, none of which have previously been identified in Staphylococcus aureus. Two independently derived S. aureus mutants having efflux-related MDR phenotypes were studied using microarray technology and a marked overexpression of an open reading frame (ORF; mepA) encoding a protein homologous with MATE family proteins was observed in both. There was concomitant overexpression of ORFs in close proximity to mepA (⬃100 bp) encoding a MarR-type regulator (mepR, upstream of mepA) and a protein of unknown function (mepB, downstream). Experiments in which mepA was overexpressed or disrupted revealed that the encoded protein has a broad substrate profile that includes several monovalent and divalent biocides and the fluoroquinolone antimicrobial agents norfloxacin and ciprofloxacin. The function of MepB is obscure, it does not contribute to the MDR phenotype conferred by MepA. MepR overexpression reversed the MDR phenotypes of both mutants by repressing mepA transcription. All three ORFs are preferentially transcribed as a single mepRAB unit, suggesting that the three genes form an operon. MDR pump was recently described for Clostridium difficile, but no member of this family has yet been identified in other gram-positive bacteria (6). Recently, our laboratory reported the in vitro selection of two S. aureus mutants that have MDR efflux phenotypes unrelated to NorA (15, 16). Herein we show that a MATE family pump is responsible for much of the MDR phenotypes exhibited by both mutants.

Multidrug efflux is an important mechanism of biocide and antimicrobial agent resistance in bacteria (18, 36). Several distinct families of membrane-based proteins mediate this process, with members of the major facilitator superfamily (MFS) constituting the largest number of described examples (4). In Staphylococcus aureus multidrug resistance (MDR) efflux proteins, hereafter referred to as MDR pumps, belonging to the MFS and small multidrug resistance families have been described previously (4, 30, 32). The best-characterized S. aureus MDR pumps are the MFS proteins QacA/B and NorA, each of which is capable of expelling multiple biocides and, with respect to NorA, selected fluoroquinolone antimicrobial agents (4, 29, 30, 32, 35). MFS and small multidrug resistance MDR pumps depend on the proton motive force for substrate transport (30). A new family of MDR pumps was recently identified in gram-negative organisms and is referred to as the multidrug and toxin extrusion (MATE) family. The first MATE MDR pump was found in Vibrio parahaemolyticus (NorM protein), and homologues have since been found in other gram-negative bacteria, plants, and the yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe (5, 5a, 11, 13, 14, 24–26, 37). Both the proton and sodium ion gradients have been identified as the energy source for substrate transport for MATE family transporters (11, 26). Analysis of S. aureus genome data in the public domain reveals many possible MDR pumps belonging to the MFS but few candidates in other transporter families. A MATE type

MATERIALS AND METHODS Bacterial strains, plasmids, media, and reagents. The strains and plasmids employed in this study are listed in Table 1. Exposure of SA-K1712 to 10-fold its norfloxacin MIC on solid media yielded SA-K1748, which, in addition to its MDR phenotype, acquired a mutation resulting in an S80F GrlA substitution (16). SA-K2068 was selected by passing S. aureus NCTC 8325-4 on a 0 to 4⫻ MIC concentration gradient of ethidium bromide (EtBr) plus moxifloxacin (15). Unless otherwise noted, all reagents were the highest grade available and were obtained from Sigma Chemical Co. (St. Louis, Mo.), or in the case of some antimicrobial agents, their respective manufacturers. Mueller-Hinton II broth was obtained from BD Biosciences, Inc., Sparks, Md., and Luria-Bertani (LB) broth from Invitrogen, Carlsbad, Calif. pK434, pK436, pK438, and pK440 were constructed by amplifying the mepR, mepA, mepB, and mepAB open reading frames, respectively, including their ribosome binding sites, from S. aureus NCTC 8325-4. PCR products were cloned into pALC2073, placing each gene under the control of a xyl/tetO tetracyclineinducible promoter (1). Plasmids were electroporated into S. aureus RN4220 and then transferred to SA-K1758, which is a norA deletion strain, by transduction using phage 85 (9, 31). The use of SA-K1758 as the host strain to characterize the MepA substrate profile was considered essential to eliminate confounding effects of NorA-mediated efflux of common substrates. All strains bearing pALC2073based constructs were grown in the presence of chloramphenicol (10 ␮g/ml) to assure plasmid maintenance and, where appropriate, were induced by 50 ng/ml of tetracycline. Strains possessing pALC2073 alone, pK434, pK436, pK438, or pK440 are indicated by the name of the parent strain appended with -P, -R, -A, -B, or -AB, respectively, to indicate that in that strain the vector only (control), mepR, mepA, mepB, or mepAB is being expressed. Genome data and bioinformatic analyses. Genome data was obtained from The Institute for Genome Research (http://www.tigr.org) (S. aureus COL), the Staphylococcus aureus Genome Sequencing Project at the University of Okla-

* Corresponding author. Mailing address: Department of Internal Medicine, Division of Infectious Diseases, Wayne State University School of Medicine, B4333 John D. Dingell VA Medical Center, 4646 John R, Detroit, MI 48201. Phone: (313) 576-4491. Fax: (313) 5761112. E-mail: [email protected]. 1857

1858

ANTIMICROB. AGENTS CHEMOTHER.

KAATZ ET AL. TABLE 1. Study strains and plasmids

Strain or plasmid

S. aureus strains NCTC 8325-4 RN4220 SA-K1712 SA-K1748 SA-K1748-P SA-K1748-R SA-K1748 mepASA-K2068 SA-K2068-P SA-K2068-R SA-K2068 mepASA-K1758 SA-K1758-P SA-K1758-A SA-K1758-B SA-K1758-AB Plasmids pALC2073 pAZ106 pK434 pK436 pK438 pK440

Relevant characteristics

Source or reference

Commonly utilized laboratory strain cured of known prophages NCTC 8325-4 r-; capable of stably maintaining recombinant plasmids NCTC 8325-4 with norA disrupted by tetK MDR mutant of SA-K1712; GrlA S80F SA-K1748 ⫹ pALC2073 SA-K1748 ⫹ pK434 mepA disrupted by fusion with lacZ MDR mutant of NCTC 8325-4 SA-K2068 ⫹ pALC2073 SA-K2068 ⫹ pK434 mepA disrupted by fusion with lacZ NCTC 8325-4 with norA deleted SA-K1758 ⫹ pALC2073 SA-K1758 ⫹ pK436 SA-K1758 ⫹ pK438 SA-K1758 ⫹ pK440

28 19 16 16 This This This 15 This This This 31 This This This This

S. aureus vector containing a tetracycline-inducible promoter controlling expression of cloned genes, Cmr Suicide vector used to disrupt mepA pALC2073-mepR pALC2073-mepA pALC2073-mepB pALC2073-mepAB

1

homa (http://www.genome.ou.edu/staph.html) (S. aureus NCTC 8325), and the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov /genomes/MICROBES/Complete.html) (S. aureus MSSA 476, MRSA 252, Mu50, N315, and MW2). Data from these sources will be referred to collectively as S. aureus genome data. Nucleotide sequence and protein hydropathy analyses were performed using DS Gene, version 1.5 (Accelrys, San Diego, Calif.). Homology analyses of protein sequences were performed using the PSI-BLAST algorithm available from the National Center for Biotechnology Information at http://www.ncbi.nlm.nih .gov/BLAST. Helix-turn-helix predictions were made using the Po ˆle BioInformatique Lyonnais software suite available at http://pbil.univ-lyon1.fr. Transcriptional profiling. The genome-wide transcriptional profiles of SAK1712, SA-K1748, S. aureus NCTC 8325-4, and SA-K2068 were determined using previously published methods (2, 7). Briefly, total RNA was extracted from strains grown in LB broth to an optical density at 550 nm of 0.4 using the RNeasy Maxi kit (QIAGEN Inc., Valencia, Calif.). The RNA was converted to cDNA using random primers. The cDNA was fragmented, biotinylated, and then hybridized with custom-designed S. aureus GeneChips (Affymetrix, Santa Clara, Calif.). Hybridizing cDNA was detected by reaction with streptavidin and comparatively quantitated and analyzed using a GeneArray scanner (Agilent Technologies, Palo Alto, Calif.) and Genespring, version 6.1, software (Silicon Genetics, Redwood City, Calif.). Antimicrobial susceptibilities. Susceptibility testing was performed in triplicate using Mueller-Hinton II broth and microdilution techniques according to NCCLS guidelines (27). EtBr efflux. The loss of EtBr from EtBr-loaded SA-K1758-P and SA-K1758-A was determined fluorometrically as described previously (16). The effect of reserpine (20 ␮g/ml) or carbonyl cyanide m-chlorophenylhydrazone (100 ␮M) on EtBr efflux also was determined. Assays were performed in triplicate, and data were compared by using the Student t test. Nucleotide sequencing and Southern blots. Nucleotide sequences were determined in both directions by the Center for Molecular Medicine and Genetics, Wayne State University, Detroit, Mich., employing an automated dideoxy chain termination-based method (33). PCR-generated mepR, mepA, and mepB probes were labeled with [␣-32P]dATP (800 Ci/mmol; NEN Life Science Products, Inc., Boston, Mass.) using the RadPrime DNA labeling system as recommended by the manufacturer (Invitrogen). Southern analyses were performed using standard techniques (34). A gyrA probe was employed for normalization purposes, and product band intensities were digitally quantitated using a phosphorimaging system (Storm

17 This This This This

study study study study study study study study study study

study study study study

860; Molecular Dynamics, Sunnyvale, Calif.) and Phoretix one-dimensional advanced software (version 2003.02; Nonlinear Dynamics Ltd., Newcastle upon Tyne, United Kingdom). Disruption of mepA. PCR was used to amplify 550 bp of internal mepA sequence near its 5⬘ end using primers incorporating BamHI and EcoRI sites. The product was cloned into pAZ106, and the resultant construct was transformed into S. aureus RN4220 (17, 19). Integration of the plasmid into the chromosome creates a transcriptional fusion between mepA and lacZ, resulting in a functional mepA knockout mutation. The mutation was transduced into SA-K1748 and SA-K2068 using phage 85. To study the effect of mepA disruption in isolation in SA-K2068, it was necessary to preserve its known mepR mutation. This was accomplished by screening mepA::lacZ transductants for their ability to turn blue on media containing 5-bromo-4-chloro-3-indolyl-␤-D-thiogalactopyranoside (X-Gal; 100 ␮g/ ml). Strains with wild-type mepR should retain a white color under these conditions because in such strains mepA expression is quite low (see Results and Discussion). One blue transductant was selected for analysis, and sequencing established that its mepR mutation was maintained. Primer extension and Northern blot analyses. Transcriptional profiling data were confirmed, and transcription start sites determined, by primer extension and Northern blot analyses. Strains were grown in LB broth and cells were collected at an optical density at 550 nm of 1. RNA was recovered using the RNeasy Midi kit (QIAGEN). For primer extension experiments, mRNAs were labeled using the primer extension system-avian myeloblastosis virus reverse transcriptase kit (Promega Corp., Madison, Wis.). The oligonucleotides used for primer extension analyses were end labeled with [␥-32P]ATP (3,000 Ci/mmol; NEN) according to procedures recommended by the manufacturer of the primer extension system, and products were visualized by autoradiography. Transcription start sites were determined by comparing the size of primer extension products with a concomitantly run dideoxy chain-termination sequencing reaction according to a protocol supplied with the primer extension system. For Northern blotting, strains were grown and RNA extracted as described above for primer extension analyses. RNA (10 ␮g) from each strain was separated in a formaldehyde-containing agarose gel, transferred to a nylon membrane, and hybridized under high-stringency conditions (42°C, 50% formamide) with the same probes as were used in Southern analyses. Autoradiography was used to detect hybridizing bands. Nucleotide sequence accession number. The GenBank accession number for the mepRAB loci of S. aureus NCTC 8325-4 is AY661734.

NOVEL S. AUREUS MATE MDR TRANSPORTER

VOL. 49, 2005

1859

TABLE 2. Susceptibilities of study strains MIC (␮g/ml) for strain: Compoundd

SA-K1712 SA-K1748 SA-K1748mepA- SA-K1748-Ra

Antimicrobial agents Norfloxacin Ciprofloxacin Moxifloxacin Monovalent cations Acriflavine BAC Cetrimide Crystal violet EtBr Pyronin Y. Rhodamine TPP TMA-DPH Divalent cations Chlorhexidine DAPI Dequalinium Hoechst 33342 Pentamidine

NCTC SA-K2068 SA-K2068mepA- SA-K2068-R SA-K1758-Pa SA-K1758-Aa 8325-4

0.31 0.08 NDc

3.13 (10)b 1.25 (16) ND

3.13 0.63 ND

3.13 0.63 ND

1.25 6.25 (5) 0.31 5.0 (16) 0.08 0.31 (4)

1.56 0.63 0.31

1.56 0.63 0.16

0.31 0.08 0.06

1.25 (4) 0.16 (2) 0.06 (1)

3.13 0.08 0.13 0.04 0.39 0.78 0.63 6.25 0.63

6.25 (2) 0.63 (8) 0.25 (2) 0.16 (4) 12.5 (32) 1.56 (2) 1.25 (2) 50 (8) 5.0 (8)

1.56 0.08 0.13 0.08 0.39 0.39 0.31 3.13 0.31

0.78 0.08 0.06 0.04 0.39 0.39 0.16 3.13 0.31

6.25 0.63 0.25 0.13 3.13 1.56 0.63 12.5 12.5

12.5 (2) 1.25 (2) 0.50 (2) 0.25 (2) 25 (8) 6.25 (4) 1.25 (2) 100 (8) 25 (2)

3.13 0.63 0.25 0.13 6.25 1.56 0.63 25.0 12.5

3.13 0.63 0.25 0.06 6.25 0.78 0.31 12.5 12.5

1.56 0.16 0.06 0.06 0.39 0.39 0.31 3.13 0.78

6.25 (4) 0.63 (4) 0.25 (4) 0.25 (4) 25.0 (64) 1.56 (4) 1.25 (4) 50.0 (16) 12.5 (16)

0.08 0.31 0.16 0.31 0.63

0.31 (4) 2.5 (8) 1.25 (8) 0.63 (2) 25 (40)

0.04 0.31 0.04 0.31 0.31

0.04 0.31 0.04 0.31 0.16

0.16 0.63 0.63 0.63 12.5

0.63 (4) 10.0 (16) 2.5 (4) 5.0 (8) 50 (4)

0.31 2.50 1.25 0.63 6.25

0.31 2.50 0.63 0.63 12.5

0.04 0.31 0.04 0.16 0.31

0.63 (16) 5.0 (16) 1.25 (32) 0.63 (4) 25.0 (80)

a

-R, -P, and -A denote the presence of pK434, pALC2073, and pK436, respectively. Values in parentheses indicate the n-fold increase in MIC compared to the appropriate parent strain. ND, not determined. d BAC, benzalkonium chloride; TPP, tetraphenylphosphonium; TMA-DPH, 1-(4-trimethylammoniumphenyl)-6-phenyl-1,3,5-hexatriene p-toluenesulfonate; DAPI, 4⬘-6-diamidino-2-phenylindole. b c

RESULTS AND DISCUSSION SA-K1748 and SA-K2068 have broad drug resistance profiles. Limited MIC data for SA-K1748 and SA-K2068 has been published previously (15, 16). In Table 2, a much broader substrate profile than was formerly appreciated for both mutants is obvious. Not included are data for compounds showing no change between parent (SA-K1712 and NCTC 8325-4) and mutant (SA-K1748 and SA-K2068) strains, which was the case for erythromycin, gentamicin, and nafcillin. SA-K1748 was derived from a norA-disrupted parent that contains a norA::tetK mutation, which rules out any role of NorA in its MDR phenotype (16). As noted previously, SAK1748 acquired a mutation in its selection process, resulting in an S80F GrlA substitution, which is certain to contribute to its raised norfloxacin MIC and may obscure MIC changes that may occur as a result of efflux pump activity. Previous work in our laboratory has shown that SA-K2068 has no known topoisomerase mutations and that its MDR phenotype is unrelated to NorA, as the sequence of norA and its promoter region are wild type and norA expression is unchanged from that of S. aureus NCTC 8325-4 (parent strain). The increased fluoroquinolone MICs in SA-K2068 are efflux related based on drug accumulation experiments performed previously (15). Of particular interest are raised MICs to both mono- and divalent cations in both mutants, including members of the quaternary ammonium (benzalkonium chloride, cetrimide, dequalinium), phenanthridine (EtBr), triphenylmethane (crystal violet), diamidine (4⬘,6-diamidino-2-phenylindole [DAPI], pentamidine) and biguanidine (chlorhexidine) families. Some of these compounds, as well as other members of these chemical families, are used as biocides/disinfectants in health care settings. This distribution of affected compounds in both mu-

tants is reminiscent of the substrate profile observed for QacA (23). Both MDR mutants have increased quantities of transcripts from three contiguous ORFs. Genome-wide transcriptional profiling data are given in Table 3. Only open reading frames (ORFs) having eightfold or greater changes in mRNA abundance are shown. Compared to the respective parent strains, the most significant changes observed for both mutants were in three contiguous ORFs corresponding to SA0404, SA0405, and SA0406 (COL numbering scheme). The corresponding ORFs for S. aureus N315 are SA0322, SA0323, and SA0324, respectively. Runoff transcripts and Northern blots (see below) verified the increased quantities of these mRNAs in both mutants, confirming transcriptional profiling data. The abundance of mRNAs of genes encoding proteins involved in carbohydrate metabolism and virulence (spa) also changed significantly in one or the other mutant strain but not in both. Thus, SA0404, SA0405, and SA0406 were considered likely candidates to be responsible for the MDR phenotypes of SA-K1748 and SA-K2068. Subsequent sequence analysis and experimental data indicated that SA0405 encodes a multidrug export protein, resulting in our naming these ORFs mepR, mepA, and mepB, respectively. Analysis of S. aureus genome data reveals that the mepRAB loci are conserved in all sequenced strains, with intergenic distances of 106 bp (mepR-A) and 103 to 106 bp (mepA-B). This genetic organization is consistent with mepRAB forming an operon. The encoded protein products of mepR and mepA (139 and 451 residues, respectively) are 98 to 100% identical between these strains. The putative mepB protein product demonstrates more variability, having a predicted size of 146 residues (S. aureus NCTC 8325, COL, MRSA 252, MSSA 476,

1860

KAATZ ET AL.

ANTIMICROB. AGENTS CHEMOTHER. TABLE 3. Genome-wide transcriptional profiling data for SA-K1748 and SA-K2068d

S. aureus COL ORF

SA0095 SA0404 SA0405 SA0406 SA0757 SA1321 SA2184 SA2185 SA2186 SA2549

Gene

spa mepR mepA mepB fruR glpD lacC lacB lacA

Fold change in expression of strain:

Product or putative function

Protein A precursor Transcriptional regulator MATE-type multidrug efflux protein Unknown function Transcriptional regulator, fructose operon Aerobic glycerol-3-phosphate dehydrogenase Tagatose-6-phosphate kinase Galactose-6-phosphate isomerase, LacB subunit Galactose-6-phosphate isomerase, LacA subunit Putative esterase

SA-K1748a

SA-K2068b

NCc 14.9 77.0 79.3 NC NC 9.3 9.5 8.2 NC

11.0 35.5 150.9 147.6 12.4 12.6 NC NC NC 8.7

Direction

Down Up Up Up Down Up Down Down Down Up

a

Compared to S. aureus SA-K1712. Compared to S. aureus NCTC 8325-4. NC, no change. d Only ORFs with changes in expression of eightfold or greater in either strain are shown, with data for both strains given for comparative purposes. b c

and MW2) or 152 residues (S. aureus N315 and Mu50). MepB homology between strains is 81 to 100%. PSI-BLAST analysis revealed that MepR has strong homology with MarR-type regulatory proteins and MepA has strong homology with MATE family transport proteins. MepR has a putative helix-turn-helix motif (residues 48 to 69), and its sequence is 25% identical and 42% similar to that of MarR. Kyte-Doolittle hydropathy analysis of MepA predicts 12 transmembrane segments, and the protein has modest homology to the MATE proteins NorM of Vibrio parahaemolyticus and CdeA of C. difficile (21 and 26% identity, respectively) (6, 26). Greater homology with MepA was observed for proteins predicted from nucleotide sequence data available for other grampositive bacteria including Listeria innocua (ORF lin0989, 41% identity), Listeria monocytogenes (ORF lmo0990, 40%), Lactococcus lactis (hypothetical protein YpbC, 38%), and another Clostridium difficile protein (EffD, 37%) (3, 8, 10; http://www .ncbi.nlm.nih.gov/genomes/MICROBES/Complete.html). All of these proteins are predicted to be MATE family transporters, but none have been characterized. MepB is predicted to be a soluble protein and has no apparent homology with proteins of known function. It is a member of COG4815, which consists of an orthologous group of uncharacterized but conserved proteins of unknown function found in the genomes of many bacteria including Listeria and Bacillus spp., Staphylococcus epidermidis, and Enterococcus faecalis. The closest homologues are found in L. innocua and L. monocytogenes (ORFs lin1093 and lmo1133, respectively; 45% identity). These ORFs are not contiguous with genes encoding MATE proteins, being more than 100 kb away in both instances. However, this arrangement does not eliminate a possible interaction between these MepB homologues and MATE proteins. MATE transporters that have been characterized include NorM and VmrA of V. parahaemolyticus, NorM of Erwinia amylovora, VcmA and VcrM of Vibrio cholerae, BexA of Bacteroides thetaiotaomicron, HmrM of Haemophilus influenzae, and most recently PmpM of Pseudomonas aeruginosa (5, 11, 13, 14, 24–26, 37). These proteins confer variable degrees of multidrug resistance with substrates including some fluoroquinolones and biocide/disinfectant agents, EtBr, acriflavine, rhoda-

mine, and occasionally aminoglycosides. Drug susceptibility data for SA-K1748 and SA-K2068 reveal raised MICs to many of the same compounds. The expression of mepRAB is augmented in both MDR mutants. Southern analyses revealed that gene amplification of mepRAB had not occurred in either mutant (data not shown). Thus, augmented mepRAB expression is the basis for the amplified signals observed for these ORFs in genome-wide transcriptional profiling. SA-K2068 has both mepR and mepA mutations. The nucleotide sequences of mepRAB of both parent strains (S. aureus NCTC 8325-4 and SA-K1712) were identical to that of S. aureus COL, as was that of SA-K1748. However, in SA-K2068, a 2-bp deletion in mepR (nucleotides 239 to 240) and a 4-bp tandem duplication in mepA (nucleotides 1299 to 1302) were found. These mutations result in the introduction of premature stop codons and predicted truncations of MepR from residues 139 to 97 and MepA from residues 451 to 439. The deleted residues of MepA (residues 440 to 451) lie in its predicted carboxy-terminal cytoplasmic tail, and their loss causes only a slight alteration in the predicted secondary structure of the protein as predicted by Kyte-Doolittle analysis. If MepA is responsible for the MDR phenotype of SA-K2068, this change has little to no impact on protein function. However, the mepR mutation in SA-K2068 is likely to result in an inactive MepR and suggests that it may be a repressor of mepAB, and perhaps mepRAB, transcription. The augmented expression of mepRAB observed in transcriptional profiling of SA-K2068 can be explained by the loss of a functional MepR and, consequently, its repressive effect. However, augmented expression of mepRAB in SA-K1748 is more difficult to explain because in that strain mepR (and, presumably, MepR) is intact. A trans-acting regulatory factor(s) must exist, the expression or activity of which is altered in SA-K1748, which is capable of overriding the presumed MepR repressive effect. A candidate protein(s) having this function was not obvious from analysis of transcriptional profiling data. There is a large inverted repeat region extending from 4 to 70 bp downstream of the mepB stop codon, which is likely to be a transcription termination signal (Fig. 1A). No other repeats were identified following the mepR or mepA coding regions.

VOL. 49, 2005

NOVEL S. AUREUS MATE MDR TRANSPORTER

1861

FIG. 1. Selected nucleotide sequences of the mepRAB locus. (A) Terminal portion of mepB and flanking DNA. The mepB stop codon is indicated, followed by inverted repeats (A and A’, B and B’, C and C’) that may function as a mepRAB transcription terminator (arrows). (B) Promoter regions of mepR and mepB. Transcription start sites (⫹1), ribosome binding sites (RBS), and start codons (SC) are shown in boldface type and highlighted with a grey background. Possible ⫺35 and ⫺10 motifs are shown in boldface type and italicized. Two transcription start sites were identified for mepB (⫹11 and ⫹ 12).

The positioning of this presumptive transcription termination signal, which has the characteristics of a typical rho-independent terminator (stem-loop followed by several U residues in an mRNA) provides additional evidence that mepR, mepA, and mepB are likely to form an operon (20). Expression of mepA verifies it as an MDR pump. The effect of expression of mepA from the heterologous tetracyclineinducible promoter of pALC2073 in a norA deletion background is illustrated in Table 2. Compared to the presence of the empty vector (SA-K1758-P), the expression of mepA (SAK1758-A) resulted in 2- to 80-fold MIC increases for all tested compounds with the exception of moxifloxacin. We have shown previously that SA-K2068 effluxes moxifloxacin, but the involved pump(s) clearly do not include MepA and were not evident in transcriptional profiling data. Independent expression of mepB had no effect on drug susceptibility, and concomitant expression of mepAB resulted in susceptibilities identical to those observed for expression of mepA alone (data not shown). These results indicate that MepB does not contribute to the MDR phenotypes of SA-K1748 or SA-K2068. SA-K1758-A exhibited a 10-fold increase in EtBr efflux compared to SA-K1758-P (P ⫽ 0.001) (Fig. 2). When reserpine (20 ␮g/ml) was included in the assay, no significant difference was observed between SA-K1758-P and SA-K1758-A (P ⫽ 0.30). Carbonyl cyanide m-chlorophenylhydrazone (100 ␮M) reduced the EtBr efflux of SA-K1758-A by 67% (data not shown). These data demonstrate the complete reserpine sensitivity of MepA as well as its dependence on intact membrane energetics for function. The best MepA substrates are pentamidine, EtBr, and the

biocide dequalinium, but reasonably good substrates also include tetraphenylphosphonium, 1-(4-trimethylammoniumphenyl)-6-phenyl-1,3,5-hexatriene ␳-toluenesulfonate, DAPI, and the clinically relevant biocide chlorhexidine. Relatively poor substrates include the fluoroquinolone antibacterial agents norfloxacin and ciprofloxacin. The small MIC increases observed for these compounds probably are not clinically relevant by themselves, but combined with other resistance mecha-

FIG. 2. MepA-mediated efflux of EtBr. Bars: 1, SA-K1758-P; 2, SA-K1758-P with reserpine (20 ␮g/ml); 3, SA-K1758-A; 4, SAK1758-A with reserpine. Results are expressed as percent reduction in fluorescence (means ⫾ standard deviations), which correlates with EtBr efflux, over 5 min.

1862

KAATZ ET AL.

nisms, they may become so. In addition, the activity of efflux pumps may lower intracellular drug concentrations to subinhibitory levels favoring the emergence of high-level resistance by way of target mutations. This hypothesis is supported by earlier work showing that inhibition of efflux pump activity in S. aureus by reserpine dramatically reduces the frequency at which mutants resistant to norfloxacin arise (22). The structural diversity between MepA substrates is considerable, which is a well-known characteristic of MDR efflux pumps. The single commonality between all of them is that each carries a positive charge or has at least one nitrogen atom that could become protonated, resulting in a positive charge. It is likely that MepA substrates interact electrostatically with the negatively charged amino acid residue(s) within the drug binding site(s). Testing this hypothesis will require crystal structure analysis of MepA in the presence and absence of substrate. MepA of SA-K2068 is fully functional. We have proposed that the minor truncation of mepA in SA-K2068 is of no functional consequence. This fact was established by cloning mepASA-K2068 into pALC2073 and expressing this construct in SA-K1758. An effect identical to that of expression of intact mepA was observed, establishing the fully functional nature of MepA in SA-K2068 (data not shown). Disruption of mepA in SA-K1748 and SA-K2068 establishes the contribution of MepA to their MDR phenotypes. Disruption of mepA resulted in a complete reversion of the MDR phenotype of SA-K1748 for all nonfluoroquinolone substrates, indicating that MepA activity is the basis for the raised MICs of all other tested compounds in this mutant (Table 2). Disruption of mepA in SA-K2068 had a similar effect, indicating that MepA is also the basis for the majority of the MDR expressed by this mutant. However, the incomplete effect of mepA disruption on DAPI resistance in SA-K2068 indicates that this mutant has a MepA-independent resistance mechanism for this compound. The responsible ORFs, which were not identified in existing microarray data, may be discovered by additional transcriptional profiling experiments of SA-K2068, its mepA-disrupted derivative, and perhaps SA-K2068 expressing pK434 (overexpressing mepR) (see below). MepR is a repressor of mepA transcription. The effect of overexpression of mepR in SA-K1748 (SA-K1748-R) and SAK2068 (SA-K2068-R) is shown in Table 2. For SA-K1748, excess MepR resulted in a reversion of all MICs except for those of norfloxacin and ciprofloxacin to values at or below those observed for the parent strain (SA-K1712). The presence of the S80F GrlA mutation in SA-K1748 likely obscures the effect that MepA, and thus also MepR, might have on its fluoroquinolone MICs. The reduction in MICs of a few compounds (acriflavine, rhodamine, dequalinium, and pentamidine) to values fourfold below those observed in SA-K1712 raises the possibility that, at least in SA-K1748, MepR may have global regulatory effects on other pumps with substrate profiles that overlap that of MepA. A comparison of the genome-wide transcriptional profile of SA-K1748 and SAK1748-R is the next logical step to test this hypothesis. With respect to SA-K2068, overexpression of mepR also reversed MIC increases of all substrates except DAPI, the MIC of which remained fourfold above that of NCTC 8325-4. The probable existence of a MepA-independent resistance mechanism for this compound has already been discussed.

ANTIMICROB. AGENTS CHEMOTHER.

FIG. 3. (A) Effect of overexpression of mepR on mepA expression. Lane 1, SA-K2068-P; lane 2, SA-K2068-R; lane 3, SA-K1748-P; lane 4, SA-K1748-R. The negative effect of MepR on mepA expression is clearly evident. (B) Expression of mepRAB. Lane 1, S. aureus NCTC 8325-4; lane 2, SA-K2068; lane 3, SA-K1712; lane 4, SA-K1748.

The MepR effect in both mutants is supportive of the conclusion that it is a negative regulator of mepA expression. The MIC-reducing effect of MepR in SA-K2068 likely results from the replacement of a dysfunctional MepR (see below) with a functional one. This presumption was tested by cloning mepRSA-K2068 into pALC2073 and determining the effect of expressing this construct in SA-K1748 and SA-K2068. No effects on MICs were observed, verifying that the truncated protein is inactive (data not shown). However, the MepR effect in SA-K1748 is intriguing in that its MepR is predicted to be intact. We presume that the abundant MepR resulting from its expression from pALC2073 overwhelms the trans-acting regulatory factor(s) stimulating mepRAB expression that we predict to exist in this strain. Northern blotting using mepR and mepA probes was performed employing RNA isolated from SA-K1748 and SAK2068 expressing either pALC2073 or pK434 (Fig. 3A). Overexpression of mepR in both mutant backgrounds essentially eliminated mepA transcripts. These data confirm MepR as a repressor of mepA transcription. Whether this effect is direct or indirect is not established by these data, but it is most likely to be direct because of the operon-like genetic organization of mepRAB. The demonstration of binding of MepR to either the mepR or mepA promoter regions would be strong evidence in favor of a direct effect. mepRAB is the favored transcript in both mutants. Northern blotting using probes for each ORF detected only mepR transcripts in parent strains (Fig. 3B). Since MepR is a repressor of mepA transcription, the detection of only mepR transcripts in parent strains is the expected result. Consistent with transcriptional profiling data and runoff transcript analyses, both mutants had marked increases in mepR and mepRAB transcripts, with small amounts of mepAB and mepB transcripts detectable. It is apparent that the favored transcript in both mutants is mepRAB, further favoring the conclusion that the mepRAB loci

NOVEL S. AUREUS MATE MDR TRANSPORTER

VOL. 49, 2005

form an operon. The well-defined hybridization signal having the highest molecular weight in each mutant is probably the intact mepRAB transcript, and the less distinct signal at a slightly lower molecular weight may consist of mepRAB degradation products. A transcript originating upstream of mepR including all or some of mepR, but none of mepA, plus degradation products of the same may explain the diffuse region of hybridization at higher molecular weights than the mepR signal in mutant strains. Independent promoters exist for mepR and mepB. Because mepA and mepB are not detectably transcribed in wild-type strains (Fig. 3B), RNA from SA-K1748 and SA-K2068 was used to identify transcription start sites for mepR, mepA, and mepB. A single start site (⫹1 nucleotide) was found for mepR, none was clearly identified for mepA, and two were apparent for mepB (⫹11 and ⫹ 12) (Fig. 1B). Multiple transcripts for a single ORF are not unusual in S. aureus; the sarA locus has three (21). Small differences in transcript size may make their resolution problematic in Northern blots, which was the case with respect to mepB. The presence of mepAB transcripts on Northern blots indicates that a mepA promoter must exist despite our inability to identify a transcription start site for this gene. The comparatively small quantities of mepAB and mepB transcripts observed in Northern blots suggest that the mepB and presumed mepA promoters are inefficient compared to that of mepR. Alternatively, compared to mepRAB transcripts, mepAB and mepB transcripts may be quite short-lived. Concluding remarks. MepA is the first S. aureus MATE family MDR transporter to be identified and is capable of transporting several clinically relevant biocides and antimicrobial agents. Our data indicate that MepR represses mepA expression and probably expression of the entire mepRAB operon based on the visualization of mepR but no mepRAB, mepAB, or mepB transcripts in the parent strains examined herein (Fig. 3B). MepB does not contribute to the function of MepA, nor does it have any susceptibility-modifying effects of its own. Its function is not clear, but its presence within the mepRAB operon suggests that it plays some role in the natural function of MepA. The purpose of this study was to characterize novel effluxrelated MDR phenotypes in S. aureus using two non-NorAmediated MDR mutants isolated at different times by independent selection methods. In both instances, parent strains were exposed to compounds that were subsequently identified as MepA substrates (i.e., norfloxacin or EtBr). This raises the possibility that substrate exposure may select for mepA upregulated mutants in clinical strains, which may have therapeutic implications. Reduced biocide susceptibility may allow S. aureus to persist in the environment, putting patients who later enter affected areas at risk of nosocomial acquisition of the organism. The contribution of efflux pumps capable of biocide or antimicrobial agent efflux to the increasing problem of MDR in clinical S. aureus strains already may be significant, and an analysis of the expression of mepA, norA, and mdeA (12) in a selection of strains possessing reserpine-reversible MDR is needed to test this hypothesis. ACKNOWLEDGMENTS This work was supported in part by VA Research Funds. We thank Ambrose Cheung for supplying pALC2073.

1863

REFERENCES 1. Bateman, B. T., N. P. Donegan, T. M. Jarry, M. Palma, and A. L. Cheung. 2001. Evaluation of a tetracycline-inducible promoter in Staphylococcus aureus in vitro and in vivo and its application in demonstrating the role of sigB in microcolony formation. Infect. Immun. 69:7851–7857. 2. Beenken, K. E., P. M. Dunman, F. McAleese, D. Macapagal, E. Murphy, S. J. Projan, J. S. Blevins, and M. S. Smeltzer. 2004. Global gene expression in Staphylococcus aureus biofilms. J. Bacteriol. 186:4665–4684. 3. Bolotin, A., P. Wincker, S. Mauger, O. Jaillon, K. Malarme, J. Weissenbach, S. Dusko Erlich, and A. Sorokin. 2001. The complete genome sequence of the lactic acid bacterium Lactococcus lactis ssp. lactis IL1403. Genome Res. 11:731–753. 4. Borgess-Walmsley, M. I., K. S. McKeegan, and A. R. Walmsley. 2003. Structure and function of efflux pumps that confer resistance to drugs. Biochem. J. 376:313–338. 5. Burse, A., H. Weingart, and M. S. Ullrich. 2004. NorM, an Erwinia amylovora multidrug efflux pump involved in in vitro competition with other epiphytic bacteria. Appl. Environ. Microbiol. 70:693–703. 5a.Diener, A. C., R. A. Gaxiola, and G. R. Fink. 2001. Arabidopsis ALF5, a multidrug efflux transporter gene family member, confers resistance to toxins. Plant Cell 13:1625–1637. 6. Dridi, L., J. Tankovic, and J.-C. Petit. 2004. CdeA of Clostridium difficile, a new multidrug efflux transporter of the MATE family. Microb. Drug Resist. 10:191–196. 7. Dunman, P. M., W. Mounts, F. McAleese, F. Immermann, D. Macapagal, E. Marsilio, L. McDougal, F. C. Tenover, P. A. Bradford, P. J. Petersen, S. J. Projan, and E. Murphy. 2004. Uses of Staphylococcus aureus GeneChips in genotyping and genetic composition analysis. J. Clin. Microbiol. 42:4275– 4283. 8. Farrow, K. A., D. Lyras, and J. I. Rood. 2001. Genomic analysis of the erythromycin resistance element Tn5398 from Clostridium difficile. Microbiology 147:2717–2728. 9. Foster, T. J. 1998. Molecular genetic analysis of staphylococcal virulence. Methods Microbiol. 27:433–454. 10. Glaser, P., L. Frangeul, C. Buchrieser, C. Rusniok, A. Amend, F. Baquero, P. Berche, H. Bloecker, P. Brandt, T. Chakraborty, A. Charbit, F. Chetouani, E. Couve´, A. de Daruvar, P. Dehoux, E. Domann, G. Dominguez-Bernal, E. Duchard, L. Durant, O. Dussurget, K.-D. Entian, H. Fsihi, F. Garcia-Del Portillo, P. Garrido, L. Gautier, W. Goebel, N. Go ´mez-Lo ´pez, T. Hain, J. Hauf, D. Jackson, L.-M. Jones, U. Kaerst, J. Kreft, M. Kuhn, F. Kunst, G. Kurapkat, E. Maduen ˜ o, A. Maitournam, J. Mata Vicente, E. Ng, H. Nedjari, G. Nordsiek, S. Novella, B. de Pablos, J.-C. Pe´rez-Diaz, R. Purcell, B. Remmel, M. Rose, T. Schlueter, N. Simoes, A. Tierrez, J.-A. Va ´zquez-Boland, H. Voss, J. Wehland, and P. Cossart. 2001. Comparative genomics of Listeria species. Science 294:849–852. 11. He, G.-X., T. Kuroda, T. Mima, Y. Morita, T. Mizushima, and T. Tsuchiya. 2004. An H⫹-coupled multidrug efflux pump, PmpM, a member of the MATE family of transporters, from Pseudomonas aeruginosa. J. Bacteriol. 186:262–265. 12. Huang, J., P. W. O’Toole, W. Shen, H. Amrine-Madsen, X. Jiang, N. Lobo, L. M. Palmer, L. Voekler, F. Fan, M. N. Gwynn, and D. McDevitt. 2004. Novel chromosomally encoded multidrug efflux transporter MdeA in Staphylococcus aureus. Antimicrob. Agents Chemother. 48:909–917. 13. Huda, M. N., J. Chen, Y. Morita, T. Kuroda, T. Mizushima, and T. Tsuchiya. 2003. Gene cloning and characterization of VcrM, a Na⫹-coupled multidrug efflux pump, from Vibrio cholerae non-O1. Microbiol. Immunol. 47:419–427. 14. Huda, M. N., Y. Morita, T. Kuroda, T. Mizushima, and T. Tsuchiya. 2001. Na⫹-driven multidrug efflux pump VcmA from Vibrio cholerae non-O1, a non-halophilic bacterium. FEMS Microbiol. Lett. 203:235–239. 15. Kaatz, G. W., V. V. Moudgal, and S. M. Seo. 2002. Identification and characterization of a novel efflux-related multidrug resistance phenotype in Staphylococcus aureus. J. Antimicrob. Chemother. 50:833–838. 16. Kaatz, G. W., S. M. Seo, L. O’Brien, M. Wahiduzzaman, and T. J. Foster. 2000. Evidence for the existence of a multidrug efflux transporter distinct from NorA in Staphylococcus aureus. Antimicrob. Agents Chemother. 44: 1404–1406. 17. Kemp, E. H., R. L. Sammons, A. Moir, D. Sun, and P. Setlow. 1991. Analysis of transcriptional control of the gerD spore germination gene of Bacillus subtilis 168. J. Bacteriol. 173:4646–4652. 18. Kim, R. B. 2002. Transporters and xenobiotic disposition. Toxicology 181182:291–297. 19. Kreiswirth, B. N., M. S. Lofdahl, M. J. Betley, M. O’Reilly, P. M. Schlievert, M. S. Bergdoll, and R. P. Novick. 1983. The toxic shock syndrome exotoxin structural gene is not detectably transmitted by prophage. Nature 305:709– 712. 20. Lewin, B. 1994. Control by RNA structure: termination and antitermination, p. 460–462. In B. Lewin (ed.), Genes V. Oxford University Press, Inc., New York, N.Y. 21. Manna, A. C., M. G. Bayer, and A. L. Cheung. 1998. Transcriptional analysis of different promoters in the sar locus of Staphylococcus aureus. J. Bacteriol. 180:3828–3836.

1864

KAATZ ET AL.

22. Markham, P. N., and A. A. Neyfakh. 1996. Inhibition of the multidrug transporter NorA prevents emergence of norfloxacin resistance in Staphylococcus aureus. Antimicrob. Agents Chemother. 40:2673–2674. 23. Mitchell, B. A., I. T. Paulsen, M. H. Brown, and R. A. Skurray. 1999. Bioenergetics of the staphylococcal multidrug export protein QacA. J. Biol. Chem. 274:3541–3548. 24. Miyamae, S., O. Ueda, F. Yoshimura, J. Hwang, Y. Tanaka, and H. Nikaido. 2001. A MATE family multidrug efflux transporter pumps out fluoroquinolones in Bacteroides thetaiotaomicron. Antimicrob. Agents Chemother. 45: 3341–3346. 25. Morita, Y., A. Kataoka, S. Shiota, T. Mizushima, and T. Tsuchiya. 2000. NorM of Vibrio parahaemolyticus is an Na⫹-driven multidrug efflux pump. J. Bacteriol. 182:6694–6697. 26. Morita, Y., K. Kodama, S. Shiota, T. Mine, A. Kataoka, T. Mizushima, and T. Tsuchiya. 1998. NorM, a putative multidrug efflux protein, of Vibrio parahaemolyticus and its homolog in Escherichia coli. Antimicrob. Agents Chemother. 42:1778–1782. 27. National Committee for Clinical Laboratory Standards. 2000. Methods for dilution susceptibility tests for bacteria that grow aerobically. Approved standard, 5th ed. NCCLS publication no. M7-A5. National Committee for Clinical Laboratory Standards, Wayne, PA. 28. Novick, R. 1967. Properties of a cryptic high-frequency transducing phage in Staphylococcus aureus. Virology 33:155–166.

ANTIMICROB. AGENTS CHEMOTHER. 29. Paulsen, I. T., M. H. Brown, T. G. Littlejohn, B. A. Mitchell, and R. A. Skurray. 1996. Multidrug resistance proteins QacA and QacB from Staphylococcus aureus: membrane topology and identification of residues involved in substrate specificity. Proc. Natl. Acad. Sci. USA 93:3630–3635. 30. Paulsen, I. T., M. H. Brown, and R. A. Skurray. 1996. Proton-dependent multidrug efflux systems. Microbiol. Rev. 60:575–608. 31. Price, C. T. D., G. W. Kaatz, and J. E. Gustafson. 2002. The multidrug efflux pump NorA is not required for salicylate-induced reduction in drug accumulation by Staphylococcus aureus. Int. J. Antimicrob. Agents 20:206–213. 32. Putman, M., H. W. van Veen, and W. N. Konings. 2000. Molecular properties of bacterial multidrug transporters. Microbiol. Mol. Biol. Rev. 64:672–693. 33. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467. 34. Southern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503–517. 35. Ubukata, K., N. Itoh-Yamashita, and M. Konno. 1989. Cloning and expression of the norA gene for fluoroquinolone resistance in Staphylococcus aureus. Antimicrob. Agents Chemother. 33:1535–1539. 36. Van Bambeke, F., E. Balzi, and P. M. Tulkens. 2000. Antibiotic efflux pumps. Biochem. Pharmacol. 60:457–470. 37. Xu, X. J., Y. Morita, T. Kuroda, T. Mizushima, and T. Tsuchiya. 2003. Molecular cloning and characterization of the HmrM multidrug efflux pump from Haemophilus influenzae Rd. Microbiol. Immunol. 47:937–943.